Reports from DownDetector suggested that the problem was global, and peaked between 9:00 and 15:00 PDT for all three services, before tailing off, though there are still residual reports of problems with all three applications.